0% found this document useful (0 votes)
4 views7 pages

Second Lecture

The document outlines the stages of analytical evolution in Big Data, which progresses from raw data to intelligent decision-making through descriptive, diagnostic, predictive, and prescriptive analytics. Each stage addresses different business questions and enhances data value, while also highlighting the challenges faced by conventional systems in handling Big Data, such as limited storage capacity, poor scalability, and inefficiencies in data analysis. The evolution emphasizes the need for advanced analytics and modern tools to effectively manage and derive insights from large and complex datasets.

Uploaded by

Priya Saini
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
4 views7 pages

Second Lecture

The document outlines the stages of analytical evolution in Big Data, which progresses from raw data to intelligent decision-making through descriptive, diagnostic, predictive, and prescriptive analytics. Each stage addresses different business questions and enhances data value, while also highlighting the challenges faced by conventional systems in handling Big Data, such as limited storage capacity, poor scalability, and inefficiencies in data analysis. The evolution emphasizes the need for advanced analytics and modern tools to effectively manage and derive insights from large and complex datasets.

Uploaded by

Priya Saini
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd

Stages of Analytical Evolution in Big Data

Analytical evolution in Big Data refers to the progression of data analysis techniques from
basic reporting to advanced decision-making systems. As organizations mature in their use of
data, analytics evolves through four main stages.

Flow of Analytical Evolution

Raw Data

Descriptive Analytics

Diagnostic Analytics

Predictive Analytics

Prescriptive Analytics

Intelligent Decision Making

Each stage answers a different business question and adds more value and intelligence.
1. Descriptive Analytics

(“What happened?”)

Description

Descriptive analytics is the first and most basic stage of data analytics. It focuses on
summarizing historical data to understand past events and performance.

It does not explain why something happened, but only describes what already occurred.

Key Characteristics

 Uses historical data


 Converts raw data into meaningful summaries
 Focuses on reporting and visualization

Techniques Used

 Data aggregation
 Statistical measures (mean, count, percentage)
 Dashboards and reports
 Charts and graphs

Examples

 Monthly sales reports


 Website traffic statistics
 Student result analysis
 Daily COVID case counts

Tools

 Excel
 SQL
 Tableau, Power BI

Outcome

Clear understanding of past performance

2. Diagnostic Analytics

(“Why did it happen?”)

Description

Diagnostic analytics goes a step further by analyzing data to identify the causes and reasons
behind past outcomes. It tries to discover relationships and dependencies in data.
Key Characteristics

 Focuses on cause-and-effect analysis


 Uses drill-down and comparison techniques
 Helps find patterns and anomalies

Techniques Used

 Root cause analysis


 Data mining
 Correlation analysis
 Drill-down and slice-and-dice analysis

Examples

 Why did sales drop last quarter?


 Why did server failures increase?
 Why did students perform poorly in a subject?

Tools

 SQL queries
 OLAP tools
 Statistical software

Outcome

Understanding reasons behind events

3. Predictive Analytics

(“What is likely to happen?”)

Description

Predictive analytics uses historical and current data along with statistical and machine
learning models to forecast future outcomes.

This stage marks the transition from analysis to prediction.

Key Characteristics

 Focuses on future trends


 Uses probability and forecasting
 Relies heavily on machine learning

Techniques Used

 Regression analysis
 Time-series forecasting
 Classification models
 Neural networks

Examples

 Predicting customer churn


 Weather forecasting
 Stock price prediction
 Disease outbreak prediction

Tools

 Python (Scikit-learn)
 R
 Hadoop, Spark
 Machine learning platforms

Outcome

Ability to anticipate future events

4. Prescriptive Analytics

(“What should we do?”)

Description

Prescriptive analytics is the most advanced stage of analytical evolution. It not only predicts
future outcomes but also recommends actions to achieve the best possible results.

It combines big data, AI, optimization, and business rules.

Key Characteristics

 Suggests optimal decisions


 Uses automation and AI
 Focuses on decision support

Techniques Used

 Optimization algorithms
 Simulation models
 Reinforcement learning
 Artificial intelligence

Examples

 Dynamic pricing in e-commerce


 Route optimization in logistics
 Personalized recommendations (Netflix, Amazon)
 Automated trading systems

Tools

 AI & ML frameworks
 Optimization solvers
 Decision management systems

Outcome

Actionable insights and intelligent decisions

Challenges of Conventional Systems

Conventional systems refer to traditional data storage, processing, and analysis systems
that were designed to handle structured, small-scale, and static data. With the rapid growth
of Big Data, these systems face several limitations.

1. Limited Data Storage Capacity

Traditional systems rely on centralized databases with fixed storage limits. They struggle to
store massive volumes of data generated from social media, sensors, IoT devices, and online
transactions.

Challenge: Cannot scale efficiently to petabytes or exabytes of data.

2. Poor Scalability

Conventional systems use vertical scaling (adding more power to a single machine), which is
expensive and limited.

Challenge: Difficult and costly to scale when data volume increases rapidly.

3. Inability to Handle Unstructured Data

Traditional databases are optimized for structured data (tables, rows, columns). Big Data
includes unstructured and semi-structured data such as:

 Images
 Videos
 Text
 Social media posts

Challenge: Conventional systems cannot process these data types effectively.

4. Slow Processing Speed

Conventional systems process data sequentially, leading to high latency.


Challenge: Cannot provide real-time or near real-time analytics

5. High Cost of Infrastructure

Upgrading traditional systems requires expensive hardware, licensed software, and


maintenance.

Challenge: Cost increases exponentially with data growth.

6. Limited Fault Tolerance

In traditional systems, failure of a single component can cause system downtime or data
loss.

Challenge: Lack of redundancy and automatic recovery mechanisms.

7. Inefficient Data Analysis

Conventional systems focus mainly on descriptive analytics and lack support for advanced
analytics such as:

 Predictive analytics
 Machine learning
 AI-based analysis

Challenge: Inability to extract deeper insights from data.

8. Data Integration Issues

Data often comes from multiple heterogeneous sources. Traditional systems find it difficult
to integrate such diverse data formats.

Challenge: Data silos and inconsistent data management.

9. Poor Handling of Velocity and Variety

Big Data is characterized by the 3Vs: Volume, Velocity, and Variety. Conventional systems
are not designed to handle:

 High-speed data streams


 Multiple data formats

Challenge: System overload and performance degradation.

10. Limited Support for Distributed Processing

Traditional systems are designed for single-node processing and lack distributed computing
capabilities.
Challenge: Cannot leverage parallel processing for faster computation.

You might also like